public_git/whisper.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/whisper.cpp.git synced 2026-10-18 04:35:33 +02:00
Code Issues Packages Projects Releases Wiki Activity

talk-llama : sync llama.cpp

Browse Source
This commit is contained in:
Georgi Gerganov
2026-06-15 10:33:53 +03:00
parent 0a3fa9ca17
commit 0ec0845110
25 changed files with 671 additions and 96 deletions
Download Patch File Download Diff File Expand all files Collapse all files
examples/talk-llama/llama-model.cpp
+16 -3
View File
@@ -287,6 +287,8 @@ static llama_model * llama_model_mapping(llm_arch arch, const llama_model_params
return new llama_model_qwen35moe(params);
case LLM_ARCH_MISTRAL3:
return new llama_model_mistral3(params);
case LLM_ARCH_EAGLE3:
return new llama_model_eagle3(params);
case LLM_ARCH_MIMO2:
return new llama_model_mimo2(params);
case LLM_ARCH_KIMI_LINEAR:
@@ -2238,7 +2240,7 @@ ggml_cgraph * llama_model::build_graph(const llm_graph_params & params) const {
// TODO: move reranking logic here and generalize
llm->build_dense_out(dense_2_out_layers, dense_2_out_layers_b, dense_3_out_layers);
llm->res->set_outputs();
llm->res->set_outputs(params);
return llm->res->get_gf();
}
@@ -2406,6 +2408,7 @@ llama_rope_type llama_model_rope_type(const llama_model * model) {
case LLM_ARCH_ERNIE4_5:
case LLM_ARCH_ERNIE4_5_MOE:
case LLM_ARCH_MISTRAL3:
case LLM_ARCH_EAGLE3:
case LLM_ARCH_MISTRAL4:
case LLM_ARCH_LLAMA_EMBED:
case LLM_ARCH_MAINCODER:
@@ -2600,8 +2603,9 @@ uint64_t llama_model_n_params(const llama_model * model) {
bool llama_model_has_encoder(const llama_model * model) {
switch (model->arch) {
case LLM_ARCH_T5: return true;
case LLM_ARCH_T5ENCODER: return true;
case LLM_ARCH_T5:
case LLM_ARCH_T5ENCODER:
case LLM_ARCH_EAGLE3: return true;
default: return false;
}
}
@@ -2687,3 +2691,12 @@ void llama_model_base::create_tensor_qkv(llama_layer & layer, int bid,
layer.wv_b = create_tensor(tn(LLM_TENSOR_ATTN_V, "bias", bid), {n_embd_v_}, TENSOR_NOT_REQUIRED);
}
}
const int32_t * llama_model_target_layer_ids(const struct llama_model * model) {
const auto & v = model->target_layer_ids;
return v.empty() ? nullptr : v.data();
}
uint32_t llama_model_target_layer_ids_n(const struct llama_model * model) {
return (uint32_t) model->target_layer_ids.size();
}